How Well Understood Are the Processes that Create Dendroclimatic Records? A Mechanistic Model of the Climatic Control on Conifer Tree-Ring Growth Dynamics - Dendroclimatology - page 66

the original dataset, which shows a similar pattern of increased growth sensitivity
to summer precipitation, and is consistent with analysis of trends in regional cli-
mate and broad-scale forcing (Anchukaitis et al. 2006 ) . Additional tree-ring data,
updated through the most recent quarter century, should show the same behavior,
if the model-based hypothesis is correct. The results point to larger-scale predictive
studies using the VS model driven by general circulation model (GCM) output.
3.8.3 Large Network Intercomparisons
Evans et al. ( 2006 ) used a continental-scale set of meteorological stations and tree-
ring chronologies to assess the skill of the Vaganov-Shashkin model across a broad
range of species and environments. They utilized a 500 km search radius around
each of 190 tree-ring chronologies to evaluate simulated chronologies derived from
the Global Historical Climatology Network (Peterson and Vose 1997 ) within that
area. This approach allowed the model to choose the station within the given search
radius which resulted in the best simulation, and presumably best reflecting climate
conditions at the actual tree-ring chronology site. This approach assumes that a sin-
gle station exists within the search radius that best approximates the mean conditions
over a given region. A similar search radius approach has been successfully used
in point-to-point regression-based reconstructions of the Palmer Drought Severity
Index (PDSI) using tree rings (Cook et al. 1999 , 2004 ) . Evans et al. ( 2006 ) found
that the model-simulated chronologies were correlated at the 95% significance level
with actual tree-ring chronologies in 176 out of 190 cases (Fig. 3.7 ) . The results
were not dependent on the size of the search radius, with similar findings for a
200 km search area (Evans et al. 2006 , results not shown). Process model skill
was about the same as that achieved for verification-period statistical modeling of
the same chronologies using robust linear multiple regression methods common in
dendroclimatology. Skillful decadal simulations were only made in a minority of
simulations either by statistical modeling or process modeling approaches, suggest-
ing that paleoclimatic interpretations of decadal climate variability from tree-ring
width data should be made with caution. Additional studies designed to identify and
remove such non-climatic biases from candidate proxy dendroclimatic datasets are
under way.
3.8.4 Uncertainties and Caveats
There are several important caveats that should accompany the interpretation of
existing simulation experiments as well as the future applications of the VS model
to new research questions in dendroclimatology. One of the strengths of the model
is that it operates on a daily timescale, and is therefore capable of accurately
simulating the response of trees to rapidly changing seasonal climate conditions
(e.g. Deslauriers and Morin 2005 ) . The daily data requirements of the model,
however, restrict both its temporal and spatial application, due the limited length and